On the tower frequency, interrupting a routine transmission seconds before the crash, someone – I presume an air carrier waiting for takeoff – blurted out a “GO AROUND!” But the die was cast.

What I find remarkable is that the aircraft's fuselage and wings were structurally relatively unscathed (notwithstanding the fire damage after the aircraft came to rest), despite the appearance of the aircraft pirouetting through 270 degrees once it slid off the runway (search: "amateur footage of San Francisco plane crash") (sure, a whole lot of luck was required for any extraneous points not to snag on the ground, thus ensuring complete disintegration). Underscores Paul's point about the structural integrity of modern airliners, and particularly the Seattle product.

Maybe you're being listened to Paul, or at least Associated Press is listening: search "plane crashes less deadly as safety improves 20130708" in theage dot com dot au

"Shocking" in its rarity? No. "Remarkable", "Fascinating" or "Impressive", maybe. It was CAVU that day with winds 6-9kts. I had filed KPAO-0Q9 and planned to transition VFR with SFO tower at 2000MSL. At the last minute my son wasn't feeling too good so I decided instead on some touch-and-goes at KPAO's 2400ft runway despite there being a dozen other planes in the pattern. When I finished, the line guy drove up and told me what transpired while I was in the air.

What's shocking is that this happened at sea level, temps were cool, the runways are huge, there's no terrain nearby, there was hardly any wind, the skies were clear and a million, it's a relatively modern advanced aircraft and there was nothing wrong with the plane, airport or ATC. If the US was statistically overdue for a fatal, I would have expected it under much different conditions.

Looking at the photo, it is obvious that the threshold, the 1000' aiming point, and all the fixed distance markers have been displaced approximately 300 feet down the runway from their original positions - judging by the painted-over black markings relative to the existing white markings. I wonder if SFO had some close calls with the sea wall in the past that resulted in creating this extra margin of safety (which clearly was not sufficent on 6 July).

It also begs the question of what would have been the result had the aircraft impacted the runway (not the seawall) at the same descent rate and groundspeed - more injuries or less?

The pilot in training was flying a visual approach. It seems the whole approach was short and slow and there was a suggestion (which I will consider unconfirmed) that the throttles were at idle but the 10,000 hour pilot pulled the nose up in the last few seconds in an attempt to go around.

I have not seen any discussion of is the visual glide slope indicators which I suspect were operating just fine at SFO. These optical indicators should have told the pilot in training and the trainer supervising his approach that it was short long before the final few seconds when there was some suggestion the approach should be aborted and a go-around performed.

I agree with everything Paul B. said about safety being very good in this industry. Still, in those cases where accidents have happened over the last decade it seems the real cause has been one stupid pilot error after another. In this one, apparently the crew flying the plane couldn't perform a visual approach and failed to reach the runway. In the Colgan and Air France accidents the crew failed to detect and correct a stall with ice involved. And the other one a plane tried to take off on the wrong runway due to poor markings and sleepy pilots.

I guess we must accept the notion that humans make mistakes and some times they are fatal. Still, it troubles me to learn the best pilots in the air (transport rated ones) still fail to perform satisfactorily in basic stick and rudder skills.

This morning's lead news story on the crash has a link to that amateur video. The NTSB now says the recorders indicate a much slower speed than the 136-knot Vref and stick shaker activation, suggesting a high sink rate. The pilot flying, the FO, reportedly has 43 hours in type.

One comment noted the lack of items which could have snagged part of the aircraft, possibly resulting in structural breakup. I have been noticing the FAA becoming really adamant about runway safety zones and break-away signs and lights, even at small GA airports, and this accident appears to show how valuable this effort is.

The ILS was Notam'd inop and therefore the autoland system was unavailable. The crew actually had to fly the airplane and there have been some pilot forum discussions lately about automated operations and resulting loss of skills by pilots.

Not only did the crew have to fly a non-ILS approach, they had to fly a visual approach. This might be common practice at certain airports (e.g. LAS) but it might be interesting to learn when the last time this international crew actually flew a visual approach was. It appears they failed to notice the visual approach slope indicators on the field.

This accident proposes an interesting question. We must assume that this pilot was IFR current. But is that sufficient for flying in a perfectly clear VFR day NOT USING ALL OF THE AUTOMATED SYSTEMS? Obviously not.

I just looked up a little information about SFO, but I am not smart enough to figure out what it means. I hope some of you big plane drivers can interpret this . . .

Runway 28L has 4 light PAPI which should have made it pretty easy to get the glide slope right on a visual approach. However, I saw a notam (which I can't really interpret) suggesting the PAPI might have been out of service - along with the glide slope. Here is a copy: "!SFO 07/046 SFO RWY 28L PAPI OTS WEF 1307062219"

A good reminder to all us GA pilots to hand fly those IFR approaches every now and again! Can't always rely on the automation or electrons.

The PAPI was NOTAM'd OTS at 2219 on the day of the accident and I suspect as a consequence of the accident, since the time is right.

It means with effect from July 6 2216 Zulu. Chairman Hersman has confirmed that the glideslope was OTS, but the localizer might not have been.

1. The commnet about not being able to use the autoland seems specious -- why would the flight crew use it in normal conditions? 2. PAPI OTS seems significant 3. But here's the sobering one... notice that the fuselage burned but there was apparently no major fire outside the fuselage. This suggests that the fire was fed by materials inside the fuselage. The same fuselage fire was observed with the CRJ that took off from the wrong (short) runway with only one survivor.

Even with the ILS glideslope out of service, a WAAS enabled GPS would have provided a glideslope to the runway as there is a published WAAS approach to 28L.

For those who haven't come across this yet, here's an excellent analysis of the approach based on flight track data:

http://flyingprofessors.net/what-happened-to-asiana-airlines-flight-214-2/

For those who haven't come across this yet, here's an excellent analysis of the approach based on flight track data:

Google "flying professors asiana" - it's the first link.

A couple of things to think about. The last B777 accident involved landing short of the runway due to ice in the fuel, resulting in a loss of power on finals (from memory). I'd be disinclined to point fingers until the investigation of this one is complete. The other thing is, notwithstanding the lack of accidents, how many close calls there have been. I cannot recall the last time Australia has had an airliner crash with hull loss but the last one I can think of was in the days of Vickers Viscounts. Yet an Airbus had a near miss in Melbourne a few years ago. The crew mis-entered the take-off weight, resulting in incorrect take-off speeds. The aircraft dragged its tail along the runway and barely cleared the perimeter fence. Australia kept it's safety record but only just. I wonder whether the US has had similar near-misses in the past few years.

Years ago (before 911), I had the opportunity to ride up front in a 747 on a flight to Heathrow. In talking with the British Airways pilots, I was surprised to learn that every landing is flown by hand and most landings are a visual approach. Also, every landing is flown by two pilots. Their reasoning was that "stuff breaks" and when the second pilot knows they have to take over control of a landing during the landing sequence, their "head is in the game".

My biggest fear with this accident is the Beuracrats and legislaters will go into "harumph mode" and decide "something needs to be done". Invariably this ends up with more uselessregulations and paperwork for what amounts to a gross error by the pilot flying. The only remedial action I could see doing any good is for a requirement during training in the simulator to perform a visual approach with no glideslope or other electronic help from the ground.

Some Airlines have rules requiring maximum use of Automation at all times. Management feels that this will enhance safety. Except; Automation is not always available due to equipment failure maintenance requirements, or other circumstances. If a Pilot is not current on actually FLYING, they may not be able to deal with the situation. We have seen too many instances where a perfectly good aircraft was lost when there was no technical reason. The 777 is a wonderful aircraft to fly; it has excellent handling, and very responsive engines.

Most airlines and fractional operators have SOPs for the Pilot Not Flying (PNF) to call out airspeed above and below Vref. Landing performance is predicated for the aircraft to be at Vref 50 feet above the threshold. My experience in this area was that we were to maintain Vref plus or minus 10 kts. If the aircraft was at -10 knots Vref the PNF was calling out Ref minus X constantly until the pilot flying acknowledged and initiated a correction to return the airspeed to Vref. Reports now indicate that the aircraft was significantly below Vref. At this point the PNF should have been talking forcefully to the pilot flying long before the tail strike on the break water.

Was the PNF calling out air speeds. We don't know at this time because the CVR recording is not available to the public. In the Buffalo crash listening to the CVR does not reveal any PNF airspeed call outs. This contributing factor was never mentioned when the train roared out of the station dictating that all crew members would have an ATP. SOP airspeed call outs were never mentioned. Having an ATP rated crew will not solve this problem.

Could this be a contributing factor in this accident?

You have to have good SOPs and use all your resources.

As Rodney stated there is great fear that those who don't have a clue will legislate actions that will not have any real impact on future safety.

Fly Safe

Why do they have 2 pilots in the cockpit? Was NO ONE watching/monitoring the approach parameters and calling out anything? If it were the FO/Low time in type flying, why didn't the Captain do anything? Lots of unanswered crew coordination questions that most likely will be answered by analysis of the flight recorders/pilot interviews.

Question: Why don't major airports have constant surveillance video on the field? Why do we have to rely on chance amateur video to help analyze incidents like this. After all, every bank, convenience store, and many cities have surveillance video cameras everywhere, and airports don't?

Wow, 43 hrs time in type....but wait, the gov't. assured us that FOs must have 1500 hrs TT, that should fix everything, right?. Somebody please ask that question to the CNF (Congress Not Flying).

As a reader noted above, the threshold has been displaced 300'. Construction is underway to add FAA mandated overrun area for 10R, not for clearance from the seawall for 28L. This is why the GS is O/S. The PAPI for 28L are displaced accordingly.

The BA 777 crash at Heathrow was due to ice in a fuel system for those specific engines. In that case, the throttles were at full but the engines weren't responding. But in this case, the throttles were allegedly at idle with no mechanical difficulties.

However, it is a testament to the design of the 777 that two similar crash landings resulted in only two fatalities.

Paul said "inanities". LOL. Gotta love them Boeing engineers.

Low and slow is not a good place to be on short final.

Actually, it seems pretty basic stuff and it reminds me of something my old CFI said many (too many) years ago: "Wanna go up, pull back on the stick. Wanna go down, pull back more." In other words, pulling back on the yoke changes the flight angle without which the plane won't climb. However, without commensurate power being added it won't be long before the plane's wings stall and it falls out from under itself. In this instance, it seems the Boeing was too low, too slow, and the pilots were too slow in recognizing the situation. Once they did figure it out it was too late. The plane nosed up just long enough for the tail to strike the rocks, then it belly-flopped onto the runway. I'll bet this results in airlines adding more LOC/ILS-out approaches to their training schedule.

Edd, you're right about that fire source. Doesn't look like it was fuel fed from outside the airplane, but actually started in the cabin or center section.

But the amateur video we showed suggests it started almost immediately after impact. If it was an electrical source, seems like it would have taken longer to propagate than it did. Perhaps some fuel got into that area somehow.

It will be interesting to follow it.

6 VFR approaches every 6 months w/o AP might be a start...

The amateur video does seem to show the rate of descent did decrease briefly just before the aircraft made its final drop to ground contact. As Jere Joiner notes, without power to replace the energy being bled off....

The fireball in the video may have been residual fuel in the center tank, ignited after a rupture during the belly-slide. If fuel exited the tank and got into the under-floor airframe that would explain the cabin being gutted.

All speculation, we'll get the details sooner than later in this case.

I agree, Mr. Richie. A pilot asked me once what percentage coupled approaches I flew. "Very few," I answered, "if the weather is good." He looked at me like I was crazy -- as though to say why I should make it hard on myself. I wanted to keep a feel for the plane when it's lower and slower. IFR approaches are different. The attention is more focused, the published missed approach procedure has to be taken into consideration, and the workload is higher. VFR flying is practice time (but don't tell the folks in back that!)

And I should have said "angle of attack" instead of "flight angle." After all, it's all in the wing. I gotta be more precise ...

This crash was not about the lack of an ILS, the slow approach speed or the 'new' Captain. It was about a failure in communication. When this event is examined from the Human Factors perspective it most likely will be determined that a: the FO was not assertive enough in stating his concerns about the deteriorating approach or b: the Captain dismissed this information. In the US we train crewmembers to be pro-active almost to the extreme, and our culture allows this. Other cultures are not so open, despite their training, and this lack of a Team mentality results in an informational isolation and even arrogance , neither of which have any place on the modern flight deck. CRM just doesn't translate sometimes.

Just curious, but has anyone talked to the UAL 747 crew which had front row seats? Luck played a major part in this crash. It could have ended up being another Teneriffe. Also the fire source appears to originate from the starboard engine on both the CNN crash videos and the tweeted photos from the passengers.

"... A pilot familiar with SFO (whose name we agreed not to publish) said the inoperative equipment has challenged many pilots who have grown accustomed to the electronics flying the approach, regardless of the weather..."

Interesting. I have a house under the area where SoCal Approach vectors arrivals for Rwy 30 approaches to KLGB. I often listen to a scanner in my office. A lot (majority?) of arrivals ask for, and get, a 'visual' approach.

"FO was not assertive enough in stating his concerns about the deteriorating approach or b: the Captain dismissed this information"

Other way around, isn't it Pete? The FO was flying, the Captain was the PNF. The cultural block was a major factor in the KAL 801 crash at Guam in 1997.

Paul, Where did you read that the FO was flying? All I have seen thus far is that the "pilot" has 10,000 hours but only 43 in type and had not been to SFO since 2004. I assumed that this was a new 777 captain. There was also some comment somewhere about this being his 9th trip and he thus lacked some sort of standing that is granted after 10. I have no idea what that is about but it was said by one of the aviators that comments for news programs.

CNN had that on the initial coverage on Saturday ir Sunday. I just looked and seems like it was a new Captain flying, not the FO. You're right.

>> there is a published WAAS approach to 28L.

Most airliners have sophisticated multi-sensor Flight Management Systems that can build a 3-degree glidepath (or any specified angle) to any point in space, however WAAS GPS typically is not one of the sensors for that system. Regular GPS is a sensor. An airline that does most of their flying outside the continental United States has little reason to install a navigation system that only works here.

"An airline that does most of their flying outside the continental United States has little reason to install a navigation system that only works here."

Also, don't op specs have to approve use of the procedures?

Building an approach on the FMS is easily done for any runway , does not require any special equipment and can be used as a reference..however it may not be approved as a primary approach (IFR)reference.

Building an approach on the FMS is easily done for any runway , does not require any special equipment and can be used as a reference..however it may not be approved as a primary approach (IFR)reference.

The news is clear that this was a visual approach. Doesn't that mean it should be flown by outside reference in preference to fancy FMS generated approach references?

This is obvious negligence on the part of airport designers and regulators to allow a fifteen foot boulder covered wall just 100 off the runway end. The wall should be replaced with a smooth ramp of sand or concrete. Isn't there some law about obstructions on runway centerline? There should be smooth grass (or water in this case) )off every runway end for at least 3000 feet or something to prevent these needless deaths.

there is noting unusual about obstructions near the beginning of a runway,,after all the touchdown zone is 3k feet down the rw

It amazes me how pilots will agree almost immediately with the press that it had to be pilot error. What are the 2 most important things pilots monitor on final. AIRSPEED & ALTITUDE !!! I don't care if it is a C-152 or a 777. You ALL need to look at the airports that Asiana flys to, before you start badmouthing the pilots. Most pilots in the U.S. would freak-out if they saw where these Asian pilots have to land. And this is not medieval feudal asia, these are high time professional pilots with over 10,000 hours in heavies, that are not more concerned with honor and respect than crashing an airliner & killing hundreds of people, including themselves ! Let's hear all ATC audio & the flight recorders IMMEDIATELY & that will tell us what REALLY happened, like maybe one of Boeings fly by wire computers failed & the engines wouldn't respond. Follow the money, who would be the big loser if that was the case. Wake up, the big companies own the FAA & NTSB; they always have. Why else would it take 12 to 18 months to play the recordings.

The left-seat pilot was flying, a captain with 10k hours but no mention yet how many years/hrs as captain and not yet checked out in type. Right seat was an instructor pilot with a little over a third of that student's flight time and new to his job, also no mention of how many years' experience. These folks fly long legs, as evidenced by the presence of four pilots on board this flight. Long legs = low landings total per flight hour? No offense intended, but i would not'a bet on these guys to win a spot landing contest against Southwest, much less a commuter airline crew. "Those red lights next to the runway are very distracting!"

"Despite the ugly outcome and even the fire, modern air transports can clearly take a beating and still remain survivablely intact and passengers, thanks to trained crew, can get out of them in a hurry."

I enjoy Avweb and Paul's writing, but what a load of churned pap.

Where's the discussion regarding the cockpit idiots who answered "No" to the flight attendent's question of whether to get the passengers out of the aircraft? It was only when the fuselage started to burn that the cabin staff started to get the passengers out.

Where's the discussion regarding the incredibly poor response time and competence of the emergency crews? Passenger's were calling 911 twenty minutes after the plane came to rest, pleading for help. It took fully half an hour for emergency crews to do anything more than watch from a safe distance. They were "afraid the plane was going to explode." When they did move in they ran over a high school girl lying on the runway.

Where's the discussion about these pathetic demonstrations of incompetency???

Maybe so.

But the cabin crew ignored the order to stay in the airplane and got the cabin evacuated. For that, they deserve credit.

We'll see what the investigation reveals about response time and effectiveness.

The first five hours of flight training for a new Student is learning the visual flight approach and landing. Every landing any pilot ever makes is visual from minimums to touchdown. What seems to be the problem here is remembering how to make the approach. Most pilots refer to it as flight to a spot landing. It is simply configuring for landing and with the trimmed indicated-airspeed there is nothing to do but maneuver the chosen spot, usually the numbers, into the center of the windscreen and keeping it there as an unmoving target by power adjustment. That is all it has ever been. We are victims of being trained to combine automation with manual flight. There cannot be both. Either use the automation or turn it off and fly the airplane...you are the Captain so be a Captain. Don't let the rule makers fly your airplane.